Persistence of Gulf of Mexico Surface Oil from the 2010 Deepwater Horizon Spill

British Petroleum (BP) began drilling the Macondo well in the Gulf of Mexico (approximately 50 m (80 km) East-Southeast of South Pass, Louisiana) using the Marianas rig in October 2009. Drilling had reached more than 9,000 feet (2,743 m) below sea level (approximately 4,000 feet (1,219 m) below the sea floor) when Hurricane Ida damaged the rig, requiring it be towed in for repairs. The Deepwater Horizon rig arrived at the Macondo site on January 31, 2010 to complete the planned additional 9,000 feet (2,743 m) of drilling (Graham et al. 2011).

After several months of drilling, a negative pressure test was begun on the morning of April 20^th as one of the final steps planned for the semisubmersible Deepwater Horizon. An unexpected loss of power, followed by an explosion occurred at approximately 9:49 p.m. that evening. The fire started by the explosion raged aboard the Deepwater Horizon oil rig on the Macondo Prospect for approximately 36 hours before the rig sank (Bly 2010). Hydrocarbons flowed from the reservoir through the well bore until it was finally capped on July 15, 2010 (Graham et al. 2011).

When the fire began on the Deepwater Horizon, no one expected it would trigger one of the largest oil spills in the Gulf of Mexico or that it would negatively impact the U.S. Gulf Coast from Texas to Florida. It was only after the rig sank that NOAA ordered acquisition of any and all imagery of the Gulf of Mexico surrounding the Macondo well. Estimates of the flow rate varied widely. Early reports by Unified Command and SkyTruth estimated only a few thousand barrels per day through the end of April 2010 (Ramseur 2011). By the end of May, estimates ranged between a few thousand barrels per day (again reported by Unified Command) to as much as 100,000 barrels per day (by Chiang as a range of 10,000 to 100,000 barrels per day). While the circumstances limit the ability to accurately determine the amount of hydrocarbons expelled from the well prior to the July 15^th capping, the On Scene Coordinator Report estimates the total discharge to be approximately 5 million barrels (210 million gallons; Ramseur 2011). Regardless of the accuracy of this value, the magnitude and length of time over which this occurred is chilling.

My work as a forensic geographer began when I left NASA in 1996 and has included analysis for one side or the other on lawsuits spawned by similar disasters, including Hurricane Katrina, railroad accidents, oil spills, marine collisions and spills, light pollution, refinery fires and explosions. Based on this experience I followed the progress of this event with some interest, expecting a call to address what would become a surprisingly difficult question to answer: "Where did the oil go?" As hydrocarbons continued to escape from the well bore at an alarming rate, the question of where the oil was and where it went became more important each day. Surprisingly, the data needed to answer this question quickly became increasingly difficult to find or to access. While NOAA had begun collecting satellite imagery from a variety of sources, including RADARSAT, ENVISAT, and MODIS, the archive of these data along with in situ data was turned over to BP within weeks of the creation of a comprehensive GIS. BP housed it behind a secure firewall. Repeated requests by others and me outside BP for access to the data were met with what appeared to be delay, misdirection, and obfuscation. While access to the source data was denied or limited, most requests were pointed to the ERMA Deepwater Gulf Response website (http://gomex.erma.noaa.gov/erma.html).

This site was developed and is maintained jointly by NOAA, EPA, and the University of New Hampshire's Coastal Response Research Center. While this site provides access to visualizations of much of the data needed to figure out where the oil went, it doesn't provide access to the daily imagery or shapefiles of oil extent created by NOAA...